Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!ames!ucbcad!ucbvax!C.ISI.EDU!Info-IBMPC From: Info-IBMPC@C.ISI.EDU.UUCP Newsgroups: mod.computers.ibm-pc Subject: Info-IBMPC Digest V6 #9 Message-ID: <8702150457.AA23121@ucbvax.Berkeley.EDU> Date: Sat, 14-Feb-87 20:28:39 EST Article-I.D.: ucbvax.8702150457.AA23121 Posted: Sat Feb 14 20:28:39 1987 Date-Received: Sun, 15-Feb-87 09:30:43 EST Sender: daemon@ucbvax.BERKELEY.EDU Reply-To: INFO-IBMPC@C.ISI.EDU Organization: The ARPA Internet Lines: 1628 Approved: info-ibmpc@b.isi.edu Info-IBMPC Digest Saturday, 14 February 1987 Volume 6 : Issue 9 This Week's Editor: Billy Brackenridge Today's Topics: LAPLACE.C SOLVE.C CONTOUR.C Address mode bug in Microsoft MASM 4.0 Multiple file handles in MS-DOS EMS Emulators may not Allow Aliasing EMS Simulators Beware Ruggedized AT Ruggedized PCs Saving Turbo Pascal Programs How to add a RESET Switch Absolute Pointers in MSC 4.0 Making COMMAND.COM Resident for MS-DOS 2.xx: Factory Simulation Software SIMFACTORY SIMSCRIPT Queuing Simulation Today's Queries: PL/I COMPILER for IBM PC Communication Program to run with 1200/75 Baud Modem (2 Msgs) More PKX34A20 info HASHTAB.C Submitted to Library UPDATE.C Revision Does THRASHER Really Work RAMdisk for EMS card (2 Msgs) PC-Write Trojan Horse (2 Msgs) Tall Tree JRAM-Disk Tall tree JRAM-Disk and Tech Service Phone Number Maxtor EXT-4380 on IBM PC/RT Memo Field Multiprocessor Architecture (2 Msgs) External 3 1/2" Drive Query SIMTEL20 Mostly Public Domain 32016 Hardware with UNIX PD32 Pascal Formatter SIMCGA and NANSI.SYS problems Music Editing Program Wanted Minix ANSI Device Driver Query Spreadsheet Templates Wanted Simple Pascal to C Translation Aid Problem with NANSI.SYS PC-286 Query PCjr. Disk Add-on Query Double Boot Problem Imagen300 Printer Driver for MS-Word Want to Run MS C From a Program Turbo Pascal Turbo Prolog Error under DR DOS+ How do you Reboot? Reading disk volume label from C or ASM DURAPAK Mass Storage EGA Information ANSI.SYS Key redefinition Limited and Help System PibAsync 1.0 Problem ---------------------------------------------------------------------- Subject: LAPLACE.C SOLVE.C CONTOUR.C Date: Thu, 12 Feb 87 21:53:47 EST From: James R. Van Zandt I'd like to contribute two programs to the library. Here are suggested entries for the catalog: SOLVE.C Solves linear systems. Based on FORTRAN program in "Computer Methods for Mathematical Computations" by Forsythe, Malcolm, and Moler. LAPLACE.C Solves Laplace's equation in 2 dimensions with arbitrary laplace.doc boundaries. Uses relaxation on a rectangular grid with plates integer arithmetic, automatic adjustments when specified boundaries don't fall on grid points. LAPLACE starts with a coarse grid, then uniformly subdivides it to improve accuracy. Input file is suitable for GRAPH, which can be used to check the boundaries, and SPLINE, which can help prepare curved boundaries. Output is suitable for CONTOUR, which will find equipotentials. PLATES is a sample data file, and PLATES.POT (see catalog entry for CONTOUR) is a sample output. Requires SOLVE.C and HSORT.C (elsewhere in the library). CONTOUR.C [This blurb got clobbered by Emacs but program is in the library. -wab] ------------------------------ Date: 28 Jan 87 19:37 -0800 From: "Ya`akov N. Miles" To: info-ibmpc@C.ISI.EDU Subject: Address mode bug in Microsoft MASM 4.0 I have noticed the following bug in Microsoft MASM version 4.0, specifically DS:1234 refers to location 1234 as it should, but [DI+DS:1234] does not cause indirect reference to location 1234, but rather is treated like [DI+1234], similarly [DX+DS:1234] causes the same problem, erroneously referring to DX + OFFSET 1234, instead of DX + LOCATION 1234 Reply: ------------------------------ Date: Tue, 10 Feb 87 08:40:39 cst From: mlw@ncsc.ARPA (Williams) To: info-ibmpc@c.isi.edu Subject: Multiple file handles in MS-DOS Had another thought about the file handle/pointer situation in MS-DOS. I haven't researched it yet, but the system's behavior makes it look as if there's only one pointer available to a file at a time. I don't know if that makes a lot of sense, but it is understandable in a single- tasking OS in a primitive sort of way. Anyhow, maybe the sopen() function in v3.x would help...it causes the file to be opened in shared mode and seems likely to support multiple file pointers. No test yet. Mark L. Williams (mlw@ncsc.arpa) ------------------------------ Date: Wed, 11 Feb 87 08:04 EST From: Frankston@MIT-MULTICS.ARPA Subject: EMS Emulators may not Allow Aliasing Be aware that the L/I/M EMS spec allows aliasing -- two windows upon the same EMS memory. You simply cannot do that with the so-called EMS emulators. ------------------------------ Date: Thu, 12 Feb 87 18:51:54 EST From: johnl@ima.ISC.COM (John R. Levine) Subject: EMS Simulators Beware A note in issue #8 suggests the use of EMS simulators for getting the effect of expanded memory when all you have is extended memory or disk, and mentions that they have performance problems. Actually, they have a worse problem -- they don't work. The software-only products simulate EMS memory by allocating a 64K bank of real memory and swapping the 16K chunks to and from extended memory (regular memory above 1MB) or the disk. These products sort of work in many cases, but the EMS spec lets you map the same page to more than one of the 16K address windows, simulators can't do that properly, so many programs do not work. I work for a company that makes a PC program that can store its data in expanded memory. We took a lot of care to make using EMS memory fast, and it turned out that since we pack as many objects as possible to a page, we'd sometimes map the same page twice when using two different objects on that page. We got a certain number of complaints about smashed data that we eventually realized were due to imperfect EMS simulation. We finally had to test the EMS at start time to see if it worked properly, and send a message to the user saying that his EMS memory seemed to be broken so we wouldn't use it. Even simulators that do work are incredibly slow since moving 16K chunks means that remap calls take milliseconds rather than microseconds (or if they're swapping to the disk, seconds rather than microseconds.) The only EMS simulator that I've seen that works is the one that comes with a Compaq 386, because it uses the 386's paging hardware to do the mapping. It's fast, too. Too bad the machine is so expensive. John Levine, ima!johnl or Levine@YALE.somethingorother PS: There are a lot of real EMS cards whose driver software didn't work properly at first, but that's another story. ------------------------------ Date: Wed, 11 Feb 87 10:15:41 PST From: Jim Anderson Subject: Ruggedized AT I just happened to see an ad it the Feb. 10, 1987 issue of PC-WEEK (p. 150) for "Industrial AT Cabinets". BTI Systems 110 Cedar Street Wellesly Hills, MA 02181 (617) 235-0319 ------------------------------ Date: Wed, 11 Feb 87 16:18:36 est From: jl42#@andrew.cmu.edu (Jay Mathew Libove) Subject: Ruggedized PCs There is a company called Texas Microsystems who sells a Rack mountable "PC" system that uses a bus-board, and has eight slots in to which you plus whatever. INCLUDING YOUR PROCESSOR CARD, such that it is easy to upgrade from one processor to the next. I have worked with these (the company I worked for last summer opted for these systems with CDC hard disks for a work environment that would include bad weather (as it were), human bumping, and 24 hours a day, 365 days a year duty cycle at 100 percent usage. They had not failed even once in the time I was there when I left to come to college. The price on the systems is also quite good- I believe that you beat IBM's price with Texas MicroSystems, and I also believe that they have a good service contract built in or available. (Or else we would not have used them.) -Jay Libove -jl42@andrew.cmu.edu -jl42@cmuccvma.bitnet ------------------------------ Date: Wed, 11 Feb 87 10:26:33 PST From: Jim Anderson Subject: Saving Turbo Pascal Programs Oops! Well, I was close in my previous statement about constants in Turbo pascal. Omar Wing also had a valid example of a situation that would cause the originally described problem. (The original problem was that running a program from Turbo Pascal gave different results than running the Turbo generated .COM file.) The bottom line is, make sure Turbo really compiles the program to create the .COM file, rather than writing a possibly "dirty" .COM file from a previously run program. Of course, if this does make a difference in the result there is something not quite legitimate happening in the program. ------------------------------ From: gatech!ucf-cs!baird@seismo.CSS.GOV (Larry A. Baird) Date: 9 Feb 87 22:39:00 GMT Subject: How to add a RESET Switch Organization: Univ. of Central Florida, Orlando The following is an article that a friend of mine got off of compuserve. I have installed the reset switch as described here ,accept I used a bigger switch that fit a plugged hole in the back of my PC. I have had no problems with this method. I also run a program out of my autoexec.bat to keep the reset switch from testing RAM. In the Feb. 87 issue of PC Tech Journal that had an article titles "Alternatives to the Big Red Switch" in which they talk about a flag that DOS used on bootup to determine if a memory test should be done. In practice CNTL-ALT-DEL sets this flag to prevent a retest on reboot. We can write a little program that sets this flag and add this program to our autoexec.bat. To add this feature do the following DEBUG set1234.com a mov ax,40 mov ds,ax mov word ptr [72],1234 int 20 r cx d w q By adding this feature our reset switch will not cause a memory test upon bootup HOW TO ADD A TRUE HARDWARE RESET BUTTON TO THE IBM PC FOR ABOUT $4.00 I ADDED A TRUE HARDWARE RESET BUTTON TO MY IBM/PC. IT'S EASY, TAKES ABOUT 30 MINUTES AND WORKS. HERE'S HOW TO DO IT. THE IBM/PC 8088 PROCESSOR DERIVES IT'S CLOCK SIGNAL FROM AN INTEL 8284A CLOCK CHIP. THIS CHIP ALSO OUTPUTS A RESET LINE. TO ACTIVATE THIS RESET LINE AN INPUT LINE CALLED POWER GOOD (PWRGOOD ON THE LOGIC DIAGRAMS) MUST BE SHORTED TO GROUND MOMENTARILY. THIS LINE COMES IN TO THE CHIP ON PIN 11. WHEN THIS PIN IS SHORTED TO GROUND AND THEN RETURNED TO NORMAL, THE 8284A GENERATES THE RESET SIGNAL ON PIN 10 WHICH IS SENT TO THE 8088 PIN 21 AND THE BOOT PROCESS BEGINS! THE 8284A IS LOCATED IN A SOCKET NEXT TO THE POWER SUPPLY | ____ ____ ___ | | | | | | | | |8088| |8087| | | | | | | | | | | POWER | | | | |___| | | | | | | SUPPLY | | | | ___ | | | | | | | | | | | | | | | | | | | | | | |____| |____| |___| | | ___ _ | | 8 | (_) | | 2 | | THIS IS IT --> | 8 | __ | | 4 | | | | |_A_| |__| | ^ |______________________________ | | CRYSTAL GO TO RADIO SHACK AND GET: (I USED THESE) (1) AN 18 PIN DIP SOCKET (CAT 276-1992) (2) A SPST MOMENTARY PUSH BUTTON SWITCH (CAT 275-1566) (3) LIGHT GAUGE WIRE (SHIELDED AUDIO 24 GA) (CAT 278-1276) SOLDER THE SWITCH ACROSS PINS 11 & 13 OF YOUR NEW SOCKET USING THE TWO INNER WIRES AS SHOWN BELOW: NOTICE THE NOTCH !! ____ ___ PIN 1 -------> | |__| | <----- PIN 18 | | | | | | SWITCH | | SOLDER ___ | PIN 13 > | __________________| |_ | | | | | | PIN 11 > | __________________| |_| | | |___| PIN 9 -------> |___________| <------ PIN 10 NOW REMOVE THE 8284A FROM THE CURRENT SOCKET AND INSERT IT IN YOUR NEW SOCKET. NOW INSERT THE NEW SOCKET CONTAINING THE 8284A IN THE OLD SOCKET ON THE MOTHERBOARD. RUN YOUR SWITCH OUT THE BACK AND NOW YOU HAVE A RESET BUTTON!! NOTE MAKE SURE THE NOTCH IN THE OLD SOCKET, NEW SOCKET, AND 8284A ARE ALL ALIGNED. DON'T SHORT ANY OTHER PINS AND YOU SHOULD HAVE NO TROUBLE. MY SYSTEM WORKS OK AND I CAN WARM BOOT AND STILL HAVE THE CONTENTS OF MY RAM DISK ON MY JRAM BOARD, EVEN IF THE SYSTEM LOCKS UP AND WON'T ALT-CNTL-DEL BOOT. IT SEEMS TO BE AS RELIABLE AS POWER OFF. GOOD LUCK DON GENTRY Larry Baird Dept. of Computer Science uucp: {ihnp4!decvax,peora}!ucf-cs!baird University of Central Florida Orlando, FL 32816 ------------------------------ Date: Wed, 11-Feb-87 09:03:30 est From: polish%voice@columbia.edu (nat polish) Subject: Absolute Pointers in MSC 4.0 Actually, it is rather trivial to get to random locations in MSC4.0. In DOS.H are two macros FP_OFF and FP_SEG which return and let you set far pointer offset and segment. Page 63 of the Library Guide. Nat Polish ------------------------------ Date: Wed, 11 Feb 87 12:34:54 EST From: weiss@linc.cis.upenn.edu (Tom Weiss) Subject: Making COMMAND.COM Resident for MS-DOS 2.xx: As far as I know, there is nothing you can put in your MS-DOS 2.11 config.sys file that will make command.com stay resident. A reasonable alternative is to install command.com 'permanently' on a ramdisk. This message is a description of an easy technique to do just that. I have two floppies (A and B) and my ramdisk is C. You may need to adjust drive specs below for your system. Since I usually boot from a:command.com, I have no shell command in config.sys. The technique is quite simple. Rename your autoexec.bat file as boot.bat. Now make a new autoexec.bat file containing the following lines: copy command.com c: copy boot.bat c:autoexec.bat c:command c:\ /p If you are using a ramdisk installed by running a .com or .exe file, add your ramdisk installation command at the top of the new autoexec.bat file. No matter what kind of ramdisk you have, be sure to enlarge it to make room for command.com and boot.bat. Now reboot your system. Command.com will be installed on your ramdisk, and should reload from there under all circumstances. How it works: When your system boots, it will load command.com normally from drive A. Then the autoexec.bat file copies command.com and your boot file to drive C (the ramdisk). The next command spawns an inferior command.com, and tells it to look for itself in C:\ on reloads. The /p tells the spawned command.com to act as if it were the top-level command.com. This has two effects: 1) It makes the spawned command.com process the exit command, so that there is no way to accidentally pop it by typing exit at dos level. 2) It makes it execute autoexec.bat. Memory cost: The memory used will be the space on your ramdisk for command.com and boot.bat, plus about 4K for spawning an inferior command.com. Comments: When the spawned command.com starts, it looks for autoexec.bat in the same place you told it to look for itself on reloads (C:\ in the example above). However, drive A is still the default drive, so you should not need to make any changes to your boot.bat file. If you desire, you can place the command 'del c:autoexec.bat' at the end of boot.bat. On my system this generates a 'batch file missing' error message at the end of the boot, but causes no difficulties. Be sure NOT to delete c:command.com. You might even want to change it to read-only. Tom Weiss weiss@eniac.seas.upenn.edu (Arpanet/CSnet) ------------------------------ Date: Wed, 11 Feb 87 10:40:14 pst From: gould9!ronb@nosc.ARPA (Ron Belanger @ CACI) To: kjs%tufts.csnet@RELAY.CS.NET Subject: Factory Simulation Software SIMFACTORY SIMSCRIPT Kevin, CACI has two products which will do the modeling you require. SIMFACTORY is a specialized product which does the type of analysis you mentioned. It requires a description of the factory layout, processing centers, processing times, flow rates etc. It takes these parameters as input and produces an animated graphics depiction of the factory operation and various textual reports without any programming effort. For a more general-purpose approach to queuing/simulation problems we have SIMSCRIPT II.5 which is the general-purpose, high-level language similar to Pascal, but containing all the built-in constructs and data structures needed for discrete and continuous simulation modeling. The PC version also does animation graphics. Since the PC version (which runs under DOS 2.0 or > ) does its own virtual memory management and multi-tasking, you won't suffer from the current DOS 640K limit while waiting (with baited breath) for DOS 5.0 (ADOS/286-DOS) to finally be released. Universities can get either package for a small handling charge. You can call us at 619-457-9681 for more info. Ron Belanger { no disclaimer here... I work for CACI! } P.S. Tufts already has PC-SIMSCRIPT. I can put you in touch with a user. A number of universities in the Boston area have mainframe versions as well. ------------------------------ Date: Thu, 12 Feb 87 10:46:33 PST From: Imants Golts Subject: Queuing Simulation Organization: Tektronix, Inc., Beaverton, OR. SLAM may be the market leader by their own accounts, but SIMAN is better. SIMAN was developed after SLAM by the same person who developed SLAM. The original SLAM was a "kludge" of the code and features of GASP IV and Q-GERT (both of which had many graduate student's hands in them). The original SIMAN was written from scratch over a period of four years by one person. SLAM will certainly give you the simulation functionality you need, but the design of SIMAN's language (somewhat GPSS-like) makes it a lot easier to develop models. I have used and taught GASP IV, SLAM, and SIMAN extensively, and I prefer SIMAN. For more information about SIMAN and CINEMA (their animation program): Systems Modeling Corporation Calder Square P.O. Box 10074 State College, PA 16805-0074 (814) 238-5919 ------------------------------ Date: Wed, 11 Feb 87 15:22 EST From: Deba Patnaik Subject: PL/I COMPILER for IBM PC Need info on any available PL/I compilers for IBM PC and compatibles. I will appreciate any pointers. --deba deba%umdc.bitnet@wiscvm.arpa ------------------------------ Date: Wed, 11 Feb 87 21:32 N From: Subject: Communication Program to run with 1200/75 Baud Modem Problem Here in Holland we have to send data via a phoneline with a modem with a standard which is different from American standard. This is done to unload the phonenetworks. The standard is more complicated and this makes a normal modem very expensive. Its normal then to use a 1200/75 baud modem because the typing speed from a normal person is less then 75 baud. This makes the modem cheaper and you keep a receiving speed from 1200 baud. The communication programs I have seen all work with the same send-receive speed. Does anyone know if there are communication programs with split send and receive speeds like 1200/75 and if it is possible with an vt100 emulator? Greetings: Martin Beekmans ------------------------------ Date: 12 Feb 1987 16:59:07 PST Subject: Communication Program to run with 1200/75 Baud Modem From: Billy You have my sympathy. Holland is a wonderful country, but deliver me from state owned and operated phone companies! I think you are out of luck unless you get non standard hardware and non standard software to go with it. The BIOS support for serial ports was poorly written. IBM tried to compensate by releasing the ASYNC Comm Package with the original PC. It also was poorly written and abandoned. Alas there is no accepted software standard for communicating with serial ports. The hardware standard is the software standard. The PC uses the 8250 chip, which was obsolete at the time the PC was designed. Don Estridge creator of the PC promised that direct screen writes and direct hardware writes to the 8250 would be supported as part of the definition of PC compatibility. You are out of luck because the 8250 chip is unable to support split baud rates. I am sure there are some chips popular in Europe that support split baud rates, but you will have to find software to go with them. If you can find communications software, you can use NANSI.SYS in the lending library to convert that software to VT-100 emulation. Last year IBM announced a software interface for serial communications that ran with the voice communications adapter and another telephone controller card. The software interface is very well designed. It allowed background communications under Topview, but was too little too late. GLASSMODEM.ASM in our lending library uses this interface. Even this interface does not allow for split baud rates. Unless the folks at Microsoft are cleverer than I think they are, there is going to be trouble with communications programs under the 286 and 386 versions of DOS. I doubt that any of the popular communications programs will run under the newer operating systems, and you can certainly bet Microsoft won't tell us anything about what is necessary to convert our programs until they and the large companies like Lotus have had a year or so head start in conversion to the "new PC standard". ------------------------------ Date: Wed, 11 Feb 87 14:05:09 est From: william E Davidsen Subject: More PKX34A20 info Recently a new version of PKARC, identified as PKX34A20, was distributed. Having gotten this version, I attempted to use it, and got an "insufficient memory" error. After many tests I concluded that there was an error in the program. When the program was run with the environment value PKARCTMP set, the program failed (in 580k free memory). The same thing happened with the values TMP, ARCTEMP, and HOME. There were no resident programs, but there may have been some interaction with device drivers. I later pulled a copy of the program from another source and found that it ran. Comparison of these versions showed that byte 407D (location 417D in debug) had a value of 08 in the version which caused problems, and 06 in the version which ran as expected. I decided to test the results compressing data with the old version (PKX33A12), the version which objected to environment values, and the version which worked correctly. Since the output of the new version is not compatible with the ARC program, I wanted to see how much I gained by using the new version. The results, given below, indicate that in some cases the compression is not as good! I was also unable to measure any improvement in speed using the new version. In the results, the tables labeled "pkarc 1.2" are the older version which produces ARC compatible files. The new version allows compatibility with the older version as an option, so I ran it both ways. The archives labeled STLOG are text logs of BBS logins, while the STBIN archives are compressions of the PKARC and PKXARC programs themselves. Note that even with the "/oc" (old compression) option set, some files were larger in the output of the new version. Conclusion: the PKX33A12 version gives speed and compatibility with the ARC program. Because of the inconsistent results obtained with version PKX34A20, I am going to wait for another release. [Tables giving running times deleted -wab] ------------------------------ Subject: HASHTAB.C Submitted to Library Date: Wed, 11 Feb 87 22:26:38 EST From: James R. Van Zandt I'm submitting the program HASHTAB.C written by Allen Holub and published in the February issue of Dr. Dobb's. HASHTAB.C Portable symbol table routines written by Allen Holub and published in the Feb 87 issue of Dr. Dobb's. Maketab() returns a pointer to a newly created symbol table (similar to fopen()). Addsym() adds a symbol, allocates an associated data area from the heap of the specified size, and returns a pointer to it. Findsym() returns the same pointer. Symbol deletion and sorted symbol printout also provided. Hash table size is specified at execution time and governs speed of access, but the table can grow gracefully to any size. [HASHTAB.C Has been added to the lending library -wab] - Jim Van Zandt (jrv@mitre-bedford) ------------------------------ Subject: UPDATE.C Revision Date: Wed, 11 Feb 87 22:24:03 EST From: James R. Van Zandt A few days ago I submitted a new version of my program UPDATE. Since then I've discovered a bug in that version near the end of the source code. Here is the single change needed to fix the bug... Unchanged: ... } envsearch("comspec",command); Changed line 431: if(find_first(command),0) /* COMMAND.COM is missing */ To: if(find_first(command,0)) /* COMMAND.COM is missing */ Unchanged: {printf("%s is missing. \n",command); fallback(); - Jim Van Zandt [UPDATE.C has been updated in the lending library. -wab] ------------------------------ Date: Thu, 12 Feb 87 07:48:41 cst From: moore@ncsc.ARPA (Moore) Subject: Does THRASHER Really Work I ran Thrasher, the program that was announced in the last digest, against my standard XT (10M drive, 3.2DOS); it determined that the optimum number of buf- fers for my system is 8 (I've had it set to either 20 or 40). Reducing the number in my CONFIG.SYS file seemed to cause a *significant* increase in the time required to boot (after POST, of course). To check this, I checked the loading time of my favorite editor (loading CONFIG.SYS): reducing buffers caused about a 33% *increase* in the editor/file loading time! Has anyone else had similar (or different?) results? Or have I once again failed to read the instructions correctly!? Jim Moore@NCSC.arpa [Dick Gillmann tried it on two very different systems (an AT and PC with external disk). Thrasher took a long time and arrived at the same conclusion (9 buffers) for both machines. From experience he feels 15 works best for his DLX bboard on a PC. Our Dbase guru here at ISI uses 15 with Dbase as well. Has anybody any different results? -wab] ------------------------------ Subject: RAMdisk for EMS card Date: Thu, 12 Feb 87 10:43:47 EST From: James R. Van Zandt I'm frustrated. I have a program that needs about 608K of user memory. I'm trying to run it on a Z150 with 640K of conventional memory and 2 MB on an AST RAMpage expanded memory card. MS-DOS plus ANSI.SYS (also required) leave enough user memory to run the program. However, adding the device drivers to give me a ramdisk on the RAMpage leaves me with too little memory. So far, I'm using DOS 2.11 (later versions are larger) and set BUFFERS=1 (which really slows things down). AST provides three device drivers - REMM.SYS to manage the expanded memory, REX.SYS to use the expanded memory to emulate extended memory, and FASTDISK.SYS to implement the ramdisk. I've set PIDS=2, which is the only parameter I know of to decrease the memory requirements, but together they still take over 15K. AST can't (won't?) help. Can anyone offer a solution? For example, does anyone have a bare-bones ramdisk program that runs on a generic EMS card? (The Intel Above Board software totals only about 5K, which would be small enough, but doesn't recognize the RAMpage.) - Jim Van Zandt (jrv@mitre-bedford) ------------------------------ Date: 13 Feb 1987 11:23:47 PST Subject: RAMdisk for EMS card From: Billy To: James R. Van Zandt You might try the Tall Tree JRAM-3 EMS card. You can run Tall Tree RAM disk, EMS memory and up to 704K dos region with the JRAM-3 card. They don't run quite the same hardware spec as AST. AST has an exclusive with desqview and desqview runs the weird AST spec, but it looks like that shouldn't be a concern. Looks like the 704K dos region will give you enough extra to make everything fit in. ------------------------------ Date: Thu, 12 Feb 87 11:12:22 EST From: "Peter J. Laughton" Subject: PC-Write Trojan Horse In light of the announcement of PC-WRITE availability to Info-IBMPC readers (volume 6, issue 8), I considered that it would be valuable to share the following warning: ----------------------------------------------------------------------- ---------------------------------------- TROJAN HORSE ALERT: BOGUS PC-WRITE 2.7x ---------------------------------------- The latest INFOWORLD (02/09/87) reports the discovery of a bogus version of PC-WRITE. Tom Wilkinson, the sysop in Los Angeles who discovered it says "the trojan version when invoked, destroys the file allo- cation table of a user's hard disk, and initiates a low level format, destroying the hard disk's data." The bad version pretends to be the latest version, PC-WRITE 2.71 and is 98,274 bytes long. The real version of 2.7 is 98,242 bytes long, and the real version of 2.71 is 98,644 bytes. Wilkinson says the version posted on Compuserve is the real version. INFOWORLD reports that "Quicksoft, PC-WRITE's developer, is offering $2500 reward for the first person who identifies the creator of the bogus program and a $5000 reward for the person who provides proof that convicts the perpetrator." Don Richardson, 02/10/87 ----------------------------------------------------------------------- Please share your experience with version 2.72. Has it run with no trouble on a hard disk system? [Here is another good reason why the INFO-IBMPC lending library accepts source files only. -wab] ------------------------------ Subject: PC-Write Date: Thu, 12 Feb 87 11:36:17 EST From: jam@mitre-bedford.ARPA According to Quicksoft, the publisher of PC-Write, the latest version is 2.71. Version 2.72 is a hack containing a booby trap, and trashes hard disks. BEWARE! Version 2.71 is a minor update of 2.7. They will not release a version 2.72. They are trying to notify bulletin boards of the existence of the bogus version, but are walking a thin line: they don't want to scare people away from PC-Write. I use version 2.7 and like it a lot. Joshua Morris jam@mitre-bedford ------------------------------ Date: Thu, 12 Feb 87 17:28:39 MEZ From: UNM406%DBNRHRZ1.BITNET@wiscvm.wisc.edu Subject: Tall Tree JRAM-Disk I have an IBM AT (512K motherboard) and recently bought a 512K Tall tree JRAM AT3-S. I installed it using page 10 for the driver (JBOOT), as page 13 is in use by a SNA 3270 emulation. (more precise: shunt P11 to 1, option s=10b for JBOOT) Now the JRAM disk is there and works, but the redefined Ctrl-Alt sequences (warm boot without loosing RAMdisk ) don't work. Our dealer seems to be completely helpless, Tall tree is far away ---- does anybody out there know what went wrong ? Thomas Miller UNM406%DBNRHRZ1.BITNET Thomas Miller 0228 73 3158 UNM406 at DBNRHRZ1 Math.Inst. Universitaet Bonn Wegelerstr.10 D - 5300 Bonn 1 (internal use only: Be3,13) ------------------------------ Date: 13 Feb 1987 11:17:56 PST Subject: Tall tree JRAM-Disk and Tech Service Phone Number From: Billy To: UNM406%DBNRHRZ1.BITNET@WISCVM.WISC.EDU The Tall Tree Tech number is constantly busy, but I got through this morning after two days of trying. We had several questions for Tall Tree so I threw yours in the queue. Tall Tree responded that your 3270 code could be stealing the interrupt or if you are running DOS 3.2 you must upgrade to version 3.80 or later of the Tall Tree JRAM software. I have used Tall Tree boards at strange addresses for a long time and never had problems. Last week I spoke with John Henderson, president of Tall Tree. He is working on a better system for handling tech calls. In the mean time the phones are busy most of the day. Tall Tree's tech info number is (415)424-8324 Their main number is (415)493-1980. ------------------------------ Date: Thu, 12 Feb 87 13:38:42 est From: kaplan@nyu.arpa (Laurence S. Kaplan) Subject: Maxtor EXT-4380 on IBM PC/RT I am currently trying to connect a Maxtor 380MB EXT-4380 to an IBM PC-RT running ASIS. The machine is using IBM's ESDI disk controller. After adding the proper configuration tables to the format program (included in their boot program) I tried to format the disk. It did not work, returning errors such as write fault and id not found to my format request. If anyone has tried to do this elsewhere, including on an AT, and has any hints or information that might help, PLEASE send me mail (I do not subscribe to this newsgroup). Laurence S. Kaplan NYU Ultracomputer Research Project 715 Broadway Rm. 1005 New York, NY 10003 (212) 460-7327 arpa: kaplan@cmcl2 uucp: {ihnp4,seismo}!cmcl2!kaplan ------------------------------ Date: Thu, 12 Feb 87 15:34:22 PST From: willis%violet.Berkeley.EDU@berkeley.edu (Willis Johnson) Subject: Memo field I've been writing C programs that access/manipulate data in dBase III+ files and would be interested in all information you can supply on file formats. I'd gladly write the memo field program mentioned in the posting to the IBM-PC news group and place the source in the public domain. Willis Johnson willis@violet.BERKELEY.EDU ------------------------------ Date: 12 Feb 87 18:24:08 EST (Thursday) Subject: Multiprocessor Architecture From: Marty I'm going to implement a dual processor system using an IBM PC (probably AT) and an 8085 with 16k dual port static ram shared with the PC. I want a good, efficient way to moderate concurrent access to data structures in dual port ram. The system constraints are such the 8085 has to run with minimal overhead, while the PC isn't as critical. Several questions regarding implementation follow: 1) The AT technical reference says the PC bus signal I/O channel ready cannot be low for more than 2.5 microseconds. What if it is? If this is a non-negotiable constraint, how can I get around it. Could I cause a restartable processor exception to somehow occur (analogous to a page fault)? 2) The Lock line on the 80286 is not used by anything on the PC board (at least it isn't on my schematics). Is there any way to get the signal onto the PC bus? Is of use in my application? 3) Is the XCHG register with memory on the PC of any use? Even though it sets the lock signal (which I don't have on the pc bus) it doesn't appear to work like a test and set. How would it be used to moderate multi-processor systems. If the lock line only lasts for one instruction cycle, my gut feel is there are problems. 4) I'd appreciate seeing any code in 8085/8086 assembler or C to implement multiprocessor semaphore primitives on dual-port RAM. Any help would be appreciated. Thanks, marty leisner xerox corporation leisner.henr@xerox.com martyl@rocksvax.uucp ------------------------------ Date: 12 Feb 1987 17:20:52 PST Subject: Multiprocessor Architecture From: Billy To: Marty Paul Mockapetris and I built a TMS320 signal processor that shared 8K bytes of static RAM in the PC I/O space. Paul designed a semaphore bit. It a trivial circuit once you have already worked out contention for the dual port RAM. Which ever processor read the bit first would get a 1. All subsequent reads would get a 0. Writing to the semaphore bit resets it to 1. It worked fine, but in practice we didn't use it much as either processor was able to interrupt the other, and that was enough to keep the shared structures from corruption. ------------------------------ Date: 13 Feb 1987 12:17:18 PST Subject: External 3 1/2" Drive Query From: Richard Gillmann I need to get an external 3 1/2" floppy disk drive for my IBM PC/AT. I'd like to hear from anyone who has done this. Where did you buy your drive? Did you use DOS 3.2 to access it? Any problems? Richard ------------------------------ Date: Fri 13 Feb 87 23:25:53-EST From: ANDERSON-RR%OSU-20@ohio-state.ARPA Subject: SIMTEL20 Could someone post a summary of how to access SIMTEL20.ARPA? Is it only available on FTP? Is there a BITNET connection (I hope so!) Can it be accessed by dial up phone lines? (I don't have FTP access.) ------------------------------ Date: Fri, 13 Feb 87 21:03:50 PST From: defron%violet.Berkeley.EDU@berkeley.edu (Daniel Efron) Subject: Mostly Public Domain 32016 Hardware with UNIX PD32 The PD32 Users' Group is announcing its latest revision of the PD32, a mostly public domain UNIX board. The board is based on the National Semi 32016 chip set. This includes the CPU, FPU, TCU, ICU and MMU. The design is 10Mhz, 2Meg. The board requires an I/O subsystem. The most supported system is an IBM PC or clone, XT, AT, etc. However the board can run (with great difficulty) on other I/O bases (especially CP/M machines). The ideal I/O system is an AT (or an XT with NEC V20) with at least a 20M hard disk. The hardware is by George Scolaro and the software (aside from UNIX) is by Dave Rand. The board is a four layer, very high quality production. The current UNIX supported is System V.2. (We are looking into Sys V.3, but no promises.) This is a full AT&T sanctioned port with all utilities and languages, but not the Writer's Workbench. The board comes complete with all the necessary I/O drivers for the PC. The UNIX uses the PC file system. It creates one big file for the UNIX file structure. This can actually be broken up into many smaller files and you can mount a floppy or a RAMDISK under the file system. We are offering the board in many configurations. In addition we will supply artwork, schematics, pal equations and prom listings to those who are interested. Also, we will supplying data books, etc. with purchases. The design is public domain, but you need permission to sell any product based on it. We are beta testing the board now and will be selling this new version in the middle of March. If you are interested these are the approximate prices. (The last two are exact.) The price does not include shipping. Bare board w/ essential parts (pals, proms, 32016 chip set) $150 Bare board w/ essential parts and UNIX Sys V.2 $550 Complete, wave-soldered kit all components, including RAM $400 Complete kit and UNIX Sys V.2 $795 Assembled and tested board and UNIX Sys V.2 $949 You can get in touch with the PD32 UG at the following address and we'll let you know more details about how to purchase. (Please send us your US-mail address via US-mail, it helps in our record keeping.) Daniel Efron 8910 Westmoreland Lane Minneapolis, MN 55426 If you have any public domain UNIX software (for SYS V) we'd like to get a copy. Please mail to me at my arpa, uucp or US-mail address. [Minix port anyone? -wab] ------------------------------ Date: Fri 13 Feb 87 23:10:33-PST From: Lee Altenberg Subject: Pascal Formatter Does anyone know of a program for the PC that takes Pascal program text and formats it in a canonical form? I have used such programs on a PDP1134, and find them quite handy. -ALTENBERG@SUMEX-AIM.ARPA. ------------------------------ Date: Wed, 11 Feb 87 11:51:07 EST From: Lee A Butler (Space Telescope|mike) Subject: SIMCGA and NANSI.SYS problems I have had problems with SIMCGA on my system. It doesn't work, but HERCBIOS.COM does. Has anyone else seen this? Further, NANSI.SYS has a problem in that it doesn't always clear the screen when told to do so. If there has been little I/O to the screen when the escape sequence to "clear-and-home" arrives it works. If the sequence is buried in the middle of a stream going to the display then there is no "clear-and-home", just a "home". This is especially annoying because MicroEmacs compiled for ANSI use sends this string along with a stream of other stuff. When you start editing the file, whatever was on the screen before is not erased before uE displays the edit buffer. Note that ANSI.SYS does not suffer from this problem. Any suggestions (especially for fixing NANSI.SYS) will be GREATLY appreciated! [Any fixes submitted will be included in our library. -wab] Lee A. Butler Space Telescope Science Institute 3700 San Martin Dr. Baltimore MD 21218 Arpanet: butler@stsci.arpa | butler@brl.arpa Usenet: seismo!stsci.arpa!butler | {noao,astrovax,cfa,charm,nrao1}!stsci!butler Phone: (301) 338-4531 ------------------------------ Date: Sat, 14 Feb 87 15:15:58 SET From: Alun Saunders Subject: Music Editing Program Wanted To: Info-IBMPC I am looking for a (preferably public-domain or shareware) program that will allow me to edit, transpose and print musical scores. Also useful would be the capability to make an attempt at playing the edited score on the IBM PC's speaker. The only examples of such programs that I have seen are either for the Mac, which are generally very powerful and also expensive (not to mention the fact that I don't have a Mac !) or are very limited in their abilities. Please reply to me personally, and I will summarize the results to the net. Thanks in advance, Alun Saunders ------------------------------ Date: Sat, 14 Feb 87 16:14 CST From: Weinstein@HI-MULTICS.ARPA Subject: Minix Can some please tell me how to get on a uni USNET mailing list called comp.os.minix. I know how to get on arpa lists but have no idea how to gateway over to USENET and get on this list. Anyone with information... pls send mail to Weinstein -at HI-MULTICS [As was mentioned in the last digest Andy Tanenbaum is the author of minix and runs the discussion group on usenet. He recommends contacting Gene Spafford spaf@gatech.edu or Brian Reid redi@decwrl.dec.com. Scott Campbell and Erone Quek have set up a minix discussion group on bitnet. The internet needs a volunteer to coordinate with these people and set up an internet distribution and archival. Applicant must have lots of time, disk space, and a good mailer as well as access to usenet. INFO-IBMPC isn't going to run any more minix messages. When someone sets up info-minix we will tell the info-ibmpc readers where they can subscribe. -wab] ------------------------------ Date: Fri, 30 Jan 87 13:33:37 EST From: Russell Nelson Subject: ANSI Device Driver Query I have recently gone back and read every issue of volume 5, and so I feel justified in asking a stupid question. To wit, what subset of the ANSI standard do 'ansi', 'bigansi', 'nansi', and 'fansi' support? Possibly a better question is, what subset of ANSI do I need to support for a Z-100 ANSI driver? -russ GEnie: BH01 BITNET:BH01@CLUTX uucp: decvax!sii!trixie!gould!clutx!bh01 [Info-HZ100@radc-tops20 is the mailing list for Z100 stuff. -rag] ------------------------------ Date: Fri, 30 Jan 1987 13:28 PST From: JGTP003@CALSTATE.BITNET Subject: Spreadsheet Templates Wanted Does anyone have any Lotus or Supercalc templates that they are willing to share? I am especially interested in ones that might be good examples for a management/public administration class, but others could prove useful as well. Thanks (in advance) Glenn Zwinger (213/498-5403) Academic Computing Coordinator California State University, Long Beach JGTP003@CALSTATE.BITNET ------------------------------ Date: Fri, 30 Jan 87 23:08:23 EST From: jcm@ORNL-MSR.ARPA (James A. Mullens) Subject: Simple Pascal to C Translation Aid This is a simple program to convert Pascal keywords and operators to C. I've found it to be valuable, though incomplete. This version is my enhancement of a program which has appeared on the Internet in the last few months. - jim mullens / jcm@ornl-msr.arpa / oak ridge national lab [PAS2C.C has been added to the Library. -rag] ------------------------------ From: uwvax!crys.wisc.edu!jojo@seismo.CSS.GOV (Jon Wesener) Date: 27 Jan 87 04:21:11 GMT Subject: Problem with NANSI.SYS I recently got the C package from MIX which includes an C compiler, editor and windowed source level debugger. Everything is really nice, especially for the price! It's just that when I run the debugger called ctrace with ansi.sys installed, the output isn't directed to the output window correctly. Supposedly, the screen should flip to the output window followed by the output being written, 'cept it's happening in the other order. Now I figured it couldn't be such any obvious error for the people at MIX to overlook so I traced the problem to my using nansi.sys instead of the ansi.sys that came with my operating system. When I use the real ansi.sys it works perfectly! I have a rom debugger which I used to trap any escape sequences that the program might be using to switch screens and that nansi.sys might be handling differently, but no escape sequences are involved. So my question to the net is, what could the difference between nansi.sys and ansi.sys be that would cause this kind of problem? jon wesener jojo@crys.wisc.edu ------------------------------ Date: Sat, 31 Jan 87 21:06:40 est From: jld#@andrew.cmu.edu (Jay Davis) To: INFO-IBMPC@C.ISI.EDU Subject: PC-286 Query Does anyone know of a good in-depth review of the new ibm pc-286? Jay Davis (andrew.cmu.edu) ------------------------------ Date: Mon, 2 Feb 87 10:12:30 EST From: "thomas" Subject: PCjr. Disk Add-on Query I am thinking about expanding my PCjr. I would like to add a second disk drive to it. Rather than buy one of the expensive drive add-on kits, I would like to do it myself and save some money. I have heard that the controller will support two drives with slight modifications, but I don't know specifically how to do it. Rather than reinvent the wheel, I thought I'd see if anyone else has already done this. Andy Thomas ------------------------------ Date: Mon, 2 Feb 87 13:21:34 est From: preedy@nswc-wo.ARPA To: info-ibmpc@c.isi.edu Subject: Double Boot Problem I have a color PC clone system with a 20 megabyte hard disk and 2 floppies. It has a hercules color graphics board and a turbo board. I have to turn the system on, off, and on again in order to get it to boot. Then it boots twice. I also have had problems with the floppy di disks not being recognized at times. Can anyone tell if I'm having compatibility problems, hard disk problems, or ???? Any suggestions will be appreciated. Pat Reedy ------------------------------ Date: Sun, 8 Feb 87 21:55:02 EST From: mm59@andrew.cmu.edu Subject: Imagen300 Printer Driver for MS-Word I would like to use MS-Word to produce output for the Imagen 300 Laser printer. Word comes with several different .PRD files but none for the Imagen. Is there anyone who has created or knows of a .PRD file to use with Word to print documents on this printer? Even a post-processor is acceptable. A printer driver for the Apple LaserWriter came with Word that creates output files in PostScript. A program that converts PostScript files into Impress files (Imagen language), if there is one of these, would work also. Please reply via E-mail or this forum. Thanks In Advance, Mark Milliman Carnegie-Mellon University Department of Electrical and Computer Engineering E-mail address: ..!ihnp4!cmucspt!ampere!mlm mlm@ampere.ece.cmu.edu ------------------------------ To: krulwich@c.cs.cmu.edu Subject: Want to Run MS C From a Program Date: Mon, 09 Feb 87 13:09:09 -0500 From: Mark Colan Bruce Krulwich asks: I am interested in having a program written in Microsoft C 4.0 generate functions on the fly (not by coincidence -- similar to LISP's compiling a LAMBDA list). The only way I have come up with is to manually decompose each function into primitives and interpret each decomposition whenever the function is executed. Is there any way to do this for real?? I suppose it would involve compiling text from a stream and linking to the current environment. Is there a way to do this?? According to Scott Treaseder of Microsoft Corp, addressing developers attending the Microsoft Windows Developers' Seminar in Boston on Jan 29, 1987, the forthcoming 286 DOS will have the ability to NOT bind a program to its subroutines until runtime [read: dynamic linking]. As I understand it, using Microsoft Windows you can get a limited form of this functionality now. Mark Colan MIT Project Athena ------------------------------ Date: 9 Feb 1987 13:14:42 EST Subject: Turbo Pascal From: Lloyd The following is from a friend of mine who is not yet on the mailing list. His comments are in response to the recent messages on this subject and concern two topics: constants and arrays. Turbo Pascal allows the use of Typed and Untyped Constants. A typed constant is a variable with a constant value. These typed "constants" may be changed via user input. A compiled .COM file containing typed constants will use the initial constant value assigned by the source code even if the user has changed the typed constant value immediately prior to creating the .COM file. Untyped constants should not be altered. They are true constants and it is good programming practice to keep them that way. Secondly, the {$R+} compiler directive should be included in all programs that use arrays until they are fully debugged to allow for range checking. Again this is good programming practice. It is well worth the sacrifice in program execution speed to let Turbo find the error instead of you. Actually, I do have a third comment. I have noticed that when a program becomes large (i.e.-approaching stack limits), Turbo will sometimes overwrite a global variable with garbage. Has anyone else experienced this problem? The only way I found to correct this problem is to minimize the # of global variables. However, I believe this to be a bug in Turbo Pascal. ------------------------------ Date: Mon, 9 Feb 87 12:16 N From: Subject: Turbo Prolog Error under DR DOS+ Hello all, I am having trouble with Turbo prolog v.1.0. I tried to run it on an Acorn Master 512 80186 10Mhz machine under DR DOS+. When I type "PROLOG" at the A> prompt; the file loads fine; but if it starts to run DOS+ reports an "Invalid Opcode Error". It then dumps some register values which I unfortunately do not have here. Do you know if Turbo prolog is indeed using invalid opcodes? Turbo Pascal v3.01 runs fine, so does Lattice C, MS Word 3.0, DbaseII and III so I think the DOS+ is okay. How can I amend Prolog/DOS+ so that Turbo Prolog runs on this system? Is there any sign in the U.S. of a Digital Research Concurrent DOS 4.1?? Thanks in advance Marc ------------------------------ Date: Mon, 9 Feb 87 17:06 EST From: Subject: How do you Reboot? I need a quick easy way to reboot a IBM PC Running PC DOS 3.1 I would like to be able to call a .com or .exe file entitled REBOOT or something along those lines. Thanks in advance, Scott Hutchinson [See the article on "how to add a reset switch" in this issue for some relevant details on this subject. -wab] ------------------------------ Date: Wed, 11 Feb 87 09:52 N From: Subject: Reading disk volume label from C or ASM Could somebody tell me how to get/set the volume label of a disk from either C or assembler. Thanks, Johan Lilius Programmer Research Unit for Computational Linguistics University of Helsinki FINLAND ------------------------------ Date: 11 February 87 15:29-GMT From: K524911%CZHRZU1A.BITNET@wiscvm.wisc.edu Subject: DURAPAK Mass Storage We are considering the buy of a DURAPAK mass storage system by SYSGEN INC., but our dealer couldn't give us detailed information because DURAPAK is new to him, too. We are thus looking for information from experienced users about access times, reliability, convenience, pro's and contra's, bugs, etc in order to have a firsthand sight of what may be our future storage system. Thanks a lot Stefan Vogel Inst. for theoretical Physics University of Zurich, Switzerland EARN address: K524911 at CZHRZU1A ------------------------------ Date: Wed, 11 Feb 1987 23:16 CST From: a.d. jensen Subject: EGA Information To: Hello... My father has recently purchased an EGA card for his PC/XT, and now he wishes to know 'all that there is to know' about it. Unfortunately, I am quite unable to fill him in. Therefore, he has asked me to look around for a 'definitive' source on the EGA -- what it does, how to program for it, or whatever, I don't know, he just said 'find it.' I am hoping that someone on the net can help me. If you know of any text, article or 'official' documentation which he wouldn't have gotten with the board, I would greatly appreciate your response (and so will he :). Please respond directly, as I do not have access to this list. Thanks. [Perhaps if he read info-ibmpc digest he would know where to find definitive EGA documentation in the hundred or so messages on the subject. Perhaps if he looked at issue 4.125 he might have learned that the IBM Seminar Proceedings V2 N11-1 contains much information on this subject. He might also have learned about some neat programs he could get for free. In the future Please don't ask for information and then admit you never read info-ibmpc. Ask for the information and then lie to me. Tell me you read info-ibmpc daily but can't remember where you saw the answer. This makes the editor feel better. -wab] a.d. jensen "Forty below keeps out the Department of Geography riff-raff." University of North Dakota - North Dakota State Motto Grand Forks, ND 58201 ------------------------------ Date: Thu, 12 Feb 87 11:11 IST From: jonathan brandon Subject: ANSI.SYS Key redefinition Limited and Help System First of all I would like to thank all those who answered my previous problem. Since the response was so fantastic I would like to ask yet another question. When using ANSI.SYS to redefine keys there seems to be a limit on the amount of characters I can feed into my redefined keys. Something around 140 I would very much appreciate any information on the subject or as an alternative maybe a cleverer way of doing this using Microsoft C. Another problem I have which is probably to complicated to solve in a short letter is that I would like to produce a LOTUS 123 type help system using the F1 key to interrupt my program in order to call upon help screens, again I would very much appreciate any help I could get again using Microsoft`s C. Many thanks in advance yours sincerely Jonathan Brandon P.S. The first problem is the more urgent of the two. [I didn't know ANSI.SYS has a key redifinition feature in it. I throw away all programs that require ANSI.SYS so I have never tried this feature. -wab] ------------------------------ Date: Thu, 12 Feb 87 16:21 N From: Subject: PibAsync 1.0 Problem Does anyone out there have experience with a public domain include file for Turbo-pascal called PibAsync.pas. I have been trying to connect to a terminal-server from an Olivetti M24 via a three-wire connection (RX, DX, Ground) using the routines in this file. I have been getting unpredictable results, which include suddenly switching to 40-character mode. Should the routines work or are there known updates around or does anyone have an alternative solution.(preferably in Turbo). Thanks, Hans Schermer Erasmus University Rotterdam Holland Hans@HROEUR5 (this is an EARN node, so you might want to add ".bitnet" ------------------------------ End of Info-IBMPC Digest ************************ -------